AI feasibility study: Optimizing crash performance using Simcenter PhysicsAI

Crush zone management is a key aspect of automotive safety design. In a collision, the crush zone absorbs much of the kinetic energy and limits intrusion into passenger areas. This whitepaper explains how to use Simcenter PhysicsAI to design rails — an essential crush‑zone component. It describes how AI models were trained on 450 finite element analysis (FEA) crash simulations with varying rail geometries and validated with an additional 50 simulations.

It also summarizes how Simcenter Hyperstudy created a workflow to optimize rail designs using PhysicsAI. Results showed strong alignment between AI and FEA outputs based on R‑squared values, visual checks and Pareto fronts. The paper concludes that Simcenter PhysicsAI models can support early rail design evaluation and optimization, delivering results up to five times faster than traditional FEA workflows.